融合多维语义表示的概率矩阵分解模型

doi:10.3969/j.issn.0372-2112.2019.09.005

电子学报 ›› 2019, Vol. 47 ›› Issue (9): 1848-1854.DOI: 10.3969/j.issn.0372-2112.2019.09.005

所属专题：优秀论文(2022)

融合多维语义表示的概率矩阵分解模型

任开旭, 王玉龙, 刘同存, 李炜

北京邮电大学网络技术研究院, 北京 100876

收稿日期:2018-10-11 修回日期:2019-04-25 出版日期:2019-09-25
- 通讯作者:
- 任开旭
- 作者简介:
- 王玉龙 男,1968年生于山东鱼台.现为北京邮电大学副教授、硕士生导师.主要研究方向为移动增值业务系统,业务网络智能化.E-mail:wangyulong@ebupt.com;刘同存 男,1984年生于山东临沂.现为北京邮电大学博士研究生.主要研究方向为推荐算法.E-mail:liutongcun@ebupt.com;李炜男,1973年出生于天津,现为北京邮电大学副教授、硕士生导师.主要研究方向为业务网络智能化,移动增值业务系统.E-mail:liwei@ebupt.com
- 基金资助:
- 国家自然科学基金 (No.61771068，No.61671079）; 北京市自然科学基金 (No.4182041）; 教育部科学技术研究重点 (重大）项目 (No.MCM20130310）

A Probabilistic Matrix Factorization Model Based on Multidimensional Semantic Representation Learning

REN Kai-xu, WANG Yu-long, LIU Tong-cun, LI Wei

Institute of Network Technology, Beijing University of Posts and Telecommunications, Beijing 100876, China

Received:2018-10-11 Revised:2019-04-25 Online:2019-09-25 Published:2019-09-25
- Corresponding author:
- REN Kai-xu
- Supported by:
- National Natural Science Foundation of China (No.61771068, No.61671079); National Natural Science Foundation of Beijing Municipality, China (No.4182041); Science and Technology Research Project (Major Project) of Ministry of Education of China (No.MCM20130310)

摘要/Abstract

摘要： 协同过滤作为推荐系统核心技术，面临严重的评分数据稀疏性问题.融合物品文本信息可以有效的解决数据稀疏性问题，然而，目前的方法侧重于提取文本的单维特征，忽略了物品语义表示的多维特性.深度挖掘物品内容的多维特性可以更加精细化描述物品的语义信息，有助于提升推荐效果.为此，本文提出基于胶囊网络的概率生成模型.模型利用胶囊网络挖掘文本的多维语义特征，并以正则化方式融入概率矩阵分解框架，建立用户与物品之间的内在关系.实验结果表明本文提出的模型具有更高的评分预测精度.

关键词: 协同过滤, 概率矩阵分解, 胶囊网络, 多维语义特征, 正则化, 混合推荐

Abstract: Collaborative filtering, as the core technology of recommendation systems, is currently facing the sparsity problem of rating data. This can be effectively solved through integrating item text information. However, current methods focus on extracting the one-dimensional features of the text, neglecting its multidimensional semantic features. Digging deeply into the multidimensional semantic features of the text can improve the recommendations. To help achieve this goal,a probabilistic matrix factorization model based on multidimensional semantic representation learning is proposed in the present study. The model uses a capsule network to mine the multidimensional semantic features of the text, and then integrates it into the probabilistic matrix decomposition framework using the regularization method to reveal hidden features linking users and items. Experimental results show that the proposed model has higher prediction accuracy.

Key words: collaborative filtering, probabilistic matrix factorization, capsule network, multidimensional semantic features, regularization, hybrid recommendation

中图分类号:

TP311

任开旭, 王玉龙, 刘同存, 等. 融合多维语义表示的概率矩阵分解模型[J]. 电子学报, 2019, 47(9): 1848-1854.

REN Kai-xu, WANG Yu-long, LIU Tong-cun, et al. A Probabilistic Matrix Factorization Model Based on Multidimensional Semantic Representation Learning[J]. Acta Electronica Sinica, 2019, 47(9): 1848-1854.

参考文献

[1] Ma H,Yang H,Lyu M R,et al.Sorec:social recommendation using probabilistic matrix factorization[A].Proceedings of the 17th ACM Conference on Information and Knowledge Management[C].California:ACM,2008.931-940.
[2] 陈克寒,韩盼盼,吴健.基于用户聚类的异构社交网络推荐算法[J].计算机学报,2013,36(2):349-359. Chen Ke Han,Han Pan Pan,Wu Jian.User clustering based social network recommendation[J].Chinese Journal of Computers,2013,36(2):349-359.(in Chinese)
[3] Melville P,Mooney R J,Nagarajan R.Content-boosted collaborative filtering for improved recommendations[A].Proceeding Eighteenth National Conference on Artificial Intelligence[C].Menlo Park:AAAI,2002.187-192.
[4] 黄贤英,熊李媛,李沁东.基于改进协同过滤算法的个性化新闻推荐技术[J].四川大学学报(自然科学版),2018,55(01):49-55. Huang X Y,Xiong L Y,Qin-Dong LI.Personalized news recommendation technology based on improved collaborative filtering algorithm[J].Journal of Sichuan University,2018,55(01):49-55.(in Chinese)
[5] Blei D M,Ng A Y,Jordan M I.Latent dirichletallocation[J].Journal of Machine Learning Research,2003,3(Jan):993-1022.
[6] Hsieh C K,Yang L,Cui Y,et al.Collaborative metric learning[A].Proceedings of the 26th International Conference on World Wide Web[C].Republic and Canton of Geneva:IW3C2,2017.193-201.
[7] Vincent P,Larochelle H,Lajoie I,et al.Stacked denoising autoencoders:learning useful representations in a deep network with a local denoising criterion[J].Journal of Machine Learning Research,2010,11(Dec):3371-3408.
[8] Kim D,Park C,Oh J,et al.Convolutional matrix factorization for document context-aware recommendation[A].Proceedings of the 10th ACM Conference on Recommender Systems[C].Boston:ACM,2016.233-240.
[9] Bansal T,Belanger D,McCallum A.Ask the gru:Multi-task learning for deep text recommendations[A].Proceedings of the 10th ACM Conference on Recommender Systems[C].Boston:ACM,2016.107-114.
[10] Li Z,Tang J.Weakly supervised deep matrix factorization for social image understanding[J].IEEE Transactions on Image Processing,2017,26(1):276-288.
[11] Li Z,Tang J,Mei T.Deep collaborative embedding for social image understanding[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2018,1(4):1-7.
[12] Liao J,Liu T,Liu M,et al.Multi-context integrated deep neural network model for next location prediction[J].IEEE Access,2018,(6):21980-21990.
[13] Sabour S,Frosst N,Hinton G E.Dynamic routing between capsules[A].Advances in Neural Information Processing Systems[C].California:NIPS,2017.3856-3866.
[14] Mnih A,Salakhutdinov R R.Probabilistic matrix factorization[A].Advances in Neural Information Processing Systems[C].California:NIPS,2008.1257-1264.
[15] Schein A I,Popescul A,Ungar L H,et al.Methods and metrics for cold-start recommendations[A].Proceedings of the 25th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval[C].Tampere:ACM,2002.253-260.
[16] Zhou Y,Wilkinson D,Schreiber R,et al.Large-scale parallel collaborative filtering for the netflix prize[A].International Conference on Algorithmic Applications in Management[C].Shanghai:AAIM,2008.337-348.
[17] Wang C,Blei D M.Collaborative topic modeling for recommending scientific articles[A].Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining[C].San Diego:ACM,2011.448-456.
[18] Wang H,Wang N,Yeung D Y.Collaborative deep learning for recommender systems[A].Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining[C].Sydney:ACM,2015.1235-1244.

融合多维语义表示的概率矩阵分解模型

A Probabilistic Matrix Factorization Model Based on Multidimensional Semantic Representation Learning

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	梁顺攀, 刘伟, 郑智中, 原福永. 基于方面和胶囊网络的跨域评分预测模型[J]. 电子学报, 2022, 50(9): 2102-2109.
[2]	韩亮, 蔡文涛, 蒲秀娟, 罗统军, 黄谦. 使用信噪比正则化LightGBM的腹部源胎儿心电信号提取方法[J]. 电子学报, 2022, 50(8): 1793-1800.
[3]	江兵兵, 何文达, 吴兴宇, 项俊浩, 洪立斌, 盛伟国. 基于自适应图学习的半监督特征选择[J]. 电子学报, 2022, 50(7): 1643-1652.
[4]	胡向东, 李之涵. 基于胶囊网络的工业互联网入侵检测方法[J]. 电子学报, 2022, 50(6): 1457-1465.
[5]	杨晓城, 杨真乙, 阎敬业, 武林, 蒋明峰. 基于拉普拉斯算子的综合孔径辐射计成像算法[J]. 电子学报, 2022, 50(2): 339-345.
[6]	刘浩阳, 林耀进, 刘景华, 吴镒潾, 毛煜, 李绍滋. 由粗到细的分层特征选择[J]. 电子学报, 2022, 50(11): 2778-2789.
[7]	张帅, 高旻, 文俊浩, 熊庆宇, 唐旭. 基于自监督学习的去流行度偏差推荐方法[J]. 电子学报, 2022, 50(10): 2361-2371.
[8]	仇祝令, 查宇飞, 吴敏, 王青. 基于注意力学习的正则化相关滤波跟踪算法[J]. 电子学报, 2020, 48(9): 1762-1768.
[9]	郭景峰, 朱晓松, 李爽. 直播电视推荐系统的评分预测算法研究[J]. 电子学报, 2020, 48(9): 1735-1740.
[10]	郑启航, 王章权, 刘半藤, 陈阳, 陈友荣. 基于加性间距胶囊网络的家庭活动识别方法研究[J]. 电子学报, 2020, 48(8): 1580-1586.
[11]	李东瑾, 杨瑞娟, 李晓柏, 董睿杰. 基于栈式稀疏降噪自编码网络的辐射源调制识别[J]. 电子学报, 2020, 48(6): 1198-1204.
[12]	吉利鹏, 倪锦根. 偏差补偿符号子带自适应滤波器[J]. 电子学报, 2020, 48(11): 2220-2225.
[13]	蒲磊, 冯新喜, 侯志强, 查宇飞, 余旺盛. 基于深度空间正则化的相关滤波跟踪算法[J]. 电子学报, 2020, 48(10): 2025-2032.
[14]	骆睿, 刘莉, 佟瑞, 李凌云. 基于Tikhonov正则化的高分辨率群时延测量与计算方法[J]. 电子学报, 2019, 47(5): 1044-1048.
[15]	李亚峰. 一种基于多字典学习的图像分割模糊方法[J]. 电子学报, 2018, 46(7): 1700-1709.